Docker Deployment

1. Prepare your environment

Install Git.
Windows / macOS: install and start Docker Desktop.
Linux: install Docker and Docker Compose.

Choose a regular directory that you can write to. On Linux, avoid system directories such as /root and /etc; we recommend a location under /opt so sandbox features are not blocked by security rules.

Open a terminal in the directory where you want to store the project (PowerShell works on Windows), then run each line:

git clone https://github.com/langbot-app/LangBot
cd LangBot/docker

Mainland China: change the image source before startup (optional)

If Docker Hub downloads are slow, edit LangBot/docker/docker-compose.yaml first. Replace every occurrence of rockchin/langbot:latest with:

docker.langbot.app/langbot-public/rockchin/langbot:latest

Save the file before starting. No other configuration changes are needed.

3. Start LangBot

On a server, use a firewall to restrict dashboard and runtime ports to trusted sources only.

In the same terminal, run:

docker compose --profile all up -d

The first startup downloads the container images. On Windows / macOS, keep Docker Desktop running.

4. Open the dashboard

Local installation: open http://localhost:5300.
Server installation: open http://YOUR_SERVER_IP:5300 and make sure the firewall allows your device to access this port.

Follow the on-screen initialization steps, then add a model and a bot.

Beta releases (optional)

To try a beta instead of the stable setup above, download the dev/4.11.x Compose file as compose.yaml in a separate writable directory. Manually change every rockchin/langbot image to rockchin/langbot:beta: Core (langbot), plugin runtime (langbot_plugin_runtime), and Box (langbot_box, when present). The beta tag is updated with every beta release. From that directory, start or update with:

docker compose --profile all pull
docker compose --profile all up -d